2.4 Field sampling
Field surveys were conducted in early August 2020. First, the active burrow entrance at each plot was estimated by the “plugging tunnels method”, in which the burrows were plugged for 3 days, and the number of plugs that were cleared by the plateau pikas to allow access to the meadow surface were recorded (Guo et al., 2012a; Yu et al., 2017b). The average number of burrow entrances with cleared plugs in 3 days was considered the density of active burrow entrances per plot; second, the area of each bare soil patch in the disturbed plot was measured by the segmentation method (Han et al., 2011), and then the sum of all bare soil patches areas in each disturbed plot was calculated as the bare soil area for that disturbed plot; third, five vegetated quadrats (1 m×1 m) were placed on the vegetated surface approximately 8 m apart along a W pattern in all plots (disturbed and undisturbed plots), and they were moved slightly to avoid bare soil patches in disturbed plots if needed; fourth, a bare soil patch was selected as a paired bare soil quadrat for each vegetated quadrat in the disturbed plot, and the distance between each paired bare soil quadrat and vegetated quadrat was as short as possible, less than approximately 1 m. Thus, there were five paired quadrats, consisting of five vegetated quadrats and five bare soil quadrats in each disturbed plot, and there were five vegetated quadrats in each undisturbed plot since this study only focused on bare soil patches induced by plateau pikas.
In each vegetated quadrat of the disturbed or undisturbed plot, all vascular plant species were identified, and their number was recorded as the plant species richness of a community. Then, all plants rooted in a quadrat were harvested into palatable and unpalatable plants (Pang & Guo, 2017). Finally, all palatable plant samples were placed into envelopes and carried back to the laboratory.
Generally, most burrows derived from plateau pika activities were less than 20 cm depth (Yu et al., 2017b), although a few burrows extended to depths of 60 cm (Fan et al., 1999). In addition, the majority of the plant roots in the alpine grasslands of the Qinghai-Tibetan Plateau are in the top 20 cm of the soil. Therefore, this study collected soil samples at a depth of 20 cm. The soil samples were collected from vegetated quadrats and bare soil quadrats for each disturbed plot, whereas they were collected from vegetated quadrats for each undisturbed plot. Before collecting the soil samples, plants and litter were cleared from the soil surface. First, a 5-cm diameter soil auger was used to collect soil samples, with which soil organic carbon and soil nutrient concentrations (total nitrogen, phosphorus and potassium) were measured; second, soil profiles in each quadrat were produced by a spade, and a stainless-steel cutting ring (the volume was 100 cm3) was used to collect soil cores to determine soil bulk density and soil water content. The soil samples for measuring soil bulk density were quickly packed into aluminum boxes with recorded weight, and each aluminum box was numbered. The aluminum boxes with fresh soil were weighed and recorded on the spot, and these aluminum boxes with soil samples were stored at 4°C and transported to the laboratory.